Run miniprot only once

[tomasbruna]

Tested on test1.sh.

Also tested on large protein inputs - "miniprothint gtf" was always matching the one miniprot would produce natively.

[tomasbruna]

Ahh, the previous implementation would not work if multiple separate protein files were on input. That's fixed.

The current implementation might still run into issues if the uptodate( [ $prot_seq_files[$i] ], [$prot_hintsfile] is triggered. I'll look into this more.

[tomasbruna]

Also, I realized IDs might clash if miniprot is run separately on multiple files (that's related to the original implementation as well). tomasbruna/miniprothint@a38f300 should take care of that.

[KatharinaHoff]

The uptodate stuff does not work in all places, anyway. I should go through the code and remove it... in the early days for BRAKER, when we had a lot of crashes, we had that idea that users would continue a failed run after fixing BRAKER... but as I said: it's buggy, either way.

The current implementation might still run into issues if the uptodate( [ $prot_seq_files[$i] ], [$prot_hintsfile] is triggered. I'll look into this more.

[KatharinaHoff]

Our HPC is going into maintenance, tomorrow morning. I will run the tests on our end before merging. So the merge might take 2-3 days from now (depending on when our HPC awakens from its upgrade).

Thank you so much for all the work that you put into this, Tomas!!! This is really good.

[tomasbruna]

The uptodate stuff does not work in all places, anyway. I should go through the code and remove it...

I was hoping you'd say that :), it seems like a big headache to maintain it.

[KatharinaHoff]

If we at some point in time fully port to python, then snakemake (or Nextflow) would take care of this... timeline for this is very unsure on our end, though. I remove uptodate from the code for now.

I was hoping you'd say that :), it seems like a big headache to maintain it.